PROJECT SUMMARY Disparities in diabetes and its complications have persisted, despite increased public health focus on reducing them. Compared to whites living with diabetes, blacks have a two-fold increased risk of stroke, amputation, and end-stage renal disease (ESRD); Hispanics and Asians have a 20 to 50% higher risk for eye problems, and at least an 80% higher risk of ESRD. Many interventions have been developed to reduce diabetes health disparities; however, despite accumulated evidence, adoption of interventions to eliminate disparities has been slow across healthcare organizations and payers. A critical barrier to developing and adopting policies to eliminate diabetes health disparities is the lack of long-term data on the economic and clinical impact of these interventions. Simulation models are mathematical representations of the complex relationships between predictors (e.g., hemoglobin A1c) and outcomes (e.g., ESRD) among specific populations. Diabetes simulation models have been used for over a decade to describe the cost-effectiveness of clinical guidelines, clinical interventions, and new drugs and devices. However, they have been rarely used to study the disparate impact between populations, i.e. health disparities. Without such a model, it is difficult to compare the relative value of different clinical and public health interventions to reduce health disparities and difficult for policymakers to decide where to allocate resources. The reason that such a model does not exist is that the majority of current diabetes simulation models rely on data from the UKPDS and other sources, which did not have large populations of Hispanic and/or Asian patients. The major challenge to developing a more useful multi-ethnic diabetes simulation model is the need to have comprehensive data across all domains of healthcare utilization, including clinically-measured risk factors (e.g., blood pressure and laboratory results) and race/ethnicity data. We will use data from Kaiser Permanente Northern California, a multi-ethnic, socioeconomically diverse population with diabetes (n~192,000) (16% Latino, 11% African American, 7% South Asian, 4% Chinese; 13% difficulty with English) to develop a mathematical model of the relationships between patient risk factors and outcomes using Kaiser data, and then to input national data and published data into the model in order to forecast the long-term implications of efforts to reduce diabetes health disparities. Our specific aims are to 1) develop a simulation model of diabetes outcomes for white, African American, Latino, South Asian, and Chinese populations; 2) forecast the impact of past changes in risk factor control on future diabetes health disparities; and 3) determine the cost-effectiveness of diabetes health disparities interventions.